Human listeners are capable of extracting an enormous amount of Information from spoken language. A single phrase may convey semantic meaning, speaker identity, and emotional state. Although hearing is a critical component of much human communication, we know surprisingly little about how the brain transforms auditory signals to accomplish these tasks. Our long-term objective is to further the development of a unified framework for understanding high-level auditory processing and how it changes with learning. Songbirds are an excellent model for the study of higher-level auditory processing. They are one of the few animal orders that learn auditory tasks of similar difficulty to humans. In addition, a number of forebrain auditory areas have been implicated in these complex recognition tasks. Including the caudal medial mesopallium (CMM). We propose to identify and map the feature selectivity of CMM neurons in adult zebra finches, using natural songs as stimuli, multiple electrodes, and a new Information theory-based technique, called maximally informative dimensions (MID). We will test the hypothesis that there is a topography of song features in CMM. We will also record responses to a set of novel songs and to the songs of the subject's father and mate. This will allow us to test how well the receptive fields derived using MID predict neural responses in general, and to examine the additional hypothesis that father's and mate's songs have a special representation (as suggested by studies of gene Induction). Our second goal is to determine how the map of feature selectivity in CMM changes with learning. We will record CMM neurons in adult zebra finches before and after behavioral discrimination training with pairs of zebra finch songs containing both shared and unique elements. Although the birds will hear all song elements while learning, only the unique components will be useful for discriminating the two songs. This will allow us to test the hypothesis that the most behaviorally relevant elements of the new songs (the unique components) will be over-represented in terms of selectivity compared to the uninformative elements. We will also use this design to examine the idea that these changes In CMM leave any selectivity for the previously learned songs (such as those of the father and mate) intact. PUBLIC HEALTH RELEVANCE: Hearing dysfunction is implicated in many disabilities including specific language impairments, dyslexia, and autism. A basic understanding of the stages of auditory processing, as well as the ways In which they change with experience, will provide a solid foundation for exploring symptoms causes of abnormal hearing, and may suggest new treatment strategies.